Intelligent control system for large-scale finish machining of agricultural products

Abstract

The invention relates to the technical field of agricultural product processing control, and discloses an intelligent control system for agricultural product large-scale finish machining, and the system comprises a general control scheduling module which is used for activating corresponding nodes to construct data routing connection; the distribution sensing module is used for constructing a material flow real-time space-time distribution data flow; the thermal control prediction module is used for generating a heat pump pre-adjusting instruction based on the energy demand; the rheological analysis module is used for inversely calculating a material rheological characteristic index by using the electrical parameters; the cooperative compensation module is used for generating a sterilization power or forming position correction value; and the beat modulation module is used for identifying the bottleneck flux and converting the bottleneck flux into a speed modulation instruction. The real-time spatial and temporal distribution density data flow of the material is constructed through the distribution sensing module, the time lag is calculated in combination with the virtual transmission queue model of the thermal control prediction module, the pre-regulation instruction for driving the heat pump unit is generated based on the energy conservation principle, and the feed-forward accurate matching of the thermal load of the drying section is realized.

Description

Technical Field

[0001] This invention relates to the field of agricultural product processing control technology, specifically to an intelligent control system for large-scale intensive processing of agricultural products. Background Technology

[0002] Modern agricultural industrialization is driving the transformation of agricultural product processing towards large-scale, continuous, and intelligent manufacturing. Agricultural product processing lines typically encompass multiple continuous physical processing stages, including pretreatment, drying, mixing, shaping, and packaging. To meet large-scale production capacity demands, these stages are generally connected via industrial conveyor belts and fieldbuses, forming an automated production system. This production model requires a control system capable of coordinating the physical equipment at each stage to ensure the continuity of material flow and the efficiency of final product output.

[0003] In existing agricultural product processing line control applications, independent unit control is commonly used. In the drying section, the control system mainly relies on temperature sensors to collect real-time temperature data within the chamber and adjusts the heat source power through a PID feedback algorithm to maintain the set value. In subsequent stages such as mixing, microwave sterilization, or solid dosage form molding, the equipment typically operates according to preset fixed process formula parameters. The overall conveying speed is generally set to a fixed rated value. The central controller is mainly responsible for the start-stop logic control of equipment in each section and the monitoring of basic operating status. Each execution unit independently completes its respective processing task according to a predetermined single logic.

[0004] However, existing agricultural product processing line control technologies suffer from lag in the response of feedback regulation to load changes in the drying stage. Random fluctuations in feed flow, coupled with the thermal inertia of heat source equipment, lead to a time mismatch between energy supply and material load, easily causing uneven drying or localized overheating. Furthermore, the lack of correlation and interaction of physical property parameters between different stages results in information silos. Raw material bulk density or moisture content drifts from batch to batch, and subsequent forming or sterilization equipment still operates with fixed parameters, making targeted corrections difficult. This leads to out-of-tolerance finished product weight or inconsistent sterilization effects. Fixed-cycle control ignores the dynamic degradation of equipment mechanical performance; when a single node experiences cylinder pressure fluctuations or torque abnormalities leading to a decline in health, the actual throughput is limited. If upstream high-speed conveying continues, material will rapidly accumulate at the bottleneck. Therefore, this invention provides an intelligent control system for large-scale refined processing of agricultural products to address the shortcomings of existing technologies. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent control system for large-scale intensive processing of agricultural products, which solves the problems of lag, batch quality fluctuations, and unplanned downtime caused by independent control of each process in existing agricultural product processing lines.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of this invention provides an intelligent control system for large-scale intensive processing of agricultural products, comprising: The overall control and scheduling module is used to parse the processing formula data of agricultural products to extract the process topology sequence, and based on the definition of the process topology sequence, selectively activate the corresponding physical nodes to build data routing connections between each process node; The distributed sensing module is used to acquire material detection signals from the pre-processing section, calculate the instantaneous linear density of the material based on the material detection signals, and construct a real-time spatiotemporal distribution density data stream of the material flow state on the conveyor belt in combination with the initial temperature of the material. The thermal control prediction module is used to construct a virtual transmission queue model using the real-time spatiotemporal distribution density data stream and the conveyor belt running speed, calculate the time lag of the material unit arriving at the drying section, and generate a pre-adjustment command to drive the heat pump unit before the material unit arrives based on the energy demand prediction. The rheological analysis module is used to monitor the mixing process in the mixing section and calculates the rheological characteristic index representing the physical properties of the dried material by inverting the electrical parameters that reflect the characteristics of the mixing load. The collaborative compensation module is used to map the rheological characteristic index to a property correction parameter, and generate a power correction value for the microwave sterilization section or a filling position correction value for the molding section based on the property correction parameter. The beat modulation module is used to calculate the real-time health status based on the operating status data, identify the bottleneck throughput capacity of the control system based on the real-time health status, and convert it into speed modulation commands adapted to each execution unit.

[0007] Preferably, the central control and scheduling module includes: The recipe parsing unit is used to output recipe data containing the process topology sequence; The routing construction unit is used to generate a topology control matrix in binary vector form based on the process topology sequence. By comparing the process topology sequence of the current task with the control system node set, it sends a closing instruction to the field execution unit in the process topology sequence to activate the corresponding node, and sends a masking instruction to the field execution unit not in the process topology sequence to lock it in standby state.

[0008] Preferably, the distributed sensing module includes: The density calculation unit is used to preprocess the image data or weight signal in the material detection signal to obtain the coverage area or instantaneous weight of the material, calculate the instantaneous linear density based on the average mass per unit area, and perform cumulative calculation on the product of the instantaneous linear density and the instantaneous running speed of the conveyor belt within a time window to generate the instantaneous mass flow rate as a feedforward prediction input.

[0009] Preferably, the thermal control prediction module includes: The virtual queue construction unit is used to maintain the virtual transmission queue. It updates the expected arrival time of the material unit to the drying section in real time by discretizing and accumulating the instantaneous running speed of the conveyor belt. The dynamic load calculation unit is used to calculate the theoretical thermal power required to maintain the target temperature of the drying section based on the principle of energy conservation. The feedforward execution unit is used to calculate the action triggering time earlier than the expected arrival time based on a preset response time window, and when the action triggering time is reached, convert the theoretical thermal power into a frequency control command to drive the heat pump unit of the drying section.

[0010] Preferably, the rheological analysis module includes: The steady-state discrimination unit is used to determine whether the mixing time has entered the preset steady-state time interval; The index calculation unit is used to synchronously collect the instantaneous drive current, instantaneous mechanical angular velocity and voltage data of the stirring motor within the steady-state time interval, calculate the ratio of the stirring motor output power to the mechanical angular velocity, perform time averaging within the steady-state interval, and calculate the rheological characteristic index to quantitatively represent the loose density and flowability of the material.

[0011] Preferably, the collaborative compensation module includes: The parameter mapping unit is used to convert the received rheological characteristic index into an estimated relative permittivity using a pre-stored nonlinear mapping curve. The power compensation unit is used to adjust the standard microwave power according to the constant energy absorption principle based on the deviation ratio of the estimated relative permittivity relative to the dielectric constant of the standard reference material, and to generate a corrected power setting value for the microwave magnetron.

[0012] Preferably, the collaborative compensation module further includes: The position servo unit is used to represent the material density change by using the difference between the rheological characteristic index and the standard rheological index benchmark value. Based on negative feedback logic, it calculates the servo correction value of the filling depth for the metering mechanism in the molding section. When the rheological characteristic index indicates that the material density is greater than the benchmark value, it drives the servo motor to reduce the filling depth to keep the weight of the finished unit constant.

[0013] Preferably, the beat modulation module includes: The health modeling unit is used to collect state characteristic data reflecting the stability of equipment operation in a sliding window manner, calculate the statistical variance of the state characteristic data within the current sliding window, and generate a normalized real-time health index based on the statistical variance. The larger the statistical variance, the lower the value of the generated real-time health index.

[0014] Preferably, the beat modulation module further includes: The bottleneck analysis unit is used to calculate the product of the rated design flux and the real-time health index of each section to obtain the maximum allowable flux, and selects the minimum value among the maximum allowable flux of all sections as the global equilibrium flux based on the principle of the weakest link effect. The synchronous execution unit is used to convert the global balanced throughput into target running speed instructions for each execution unit according to a preset process transfer ratio coefficient.

[0015] A second aspect of the present invention also provides an intelligent control method for large-scale intensive processing of agricultural products, comprising the following steps: S1. Parse the processing formula data to lock the target process topology sequence, and according to the definition logic of the process topology sequence, issue control commands to the fieldbus to activate specific field execution units, thus building a bottom-level data interaction path that runs through the entire line. S2. Obtain the material detection signal from the pre-processing section, convert it into a real-time spatiotemporal distribution density data stream reflecting the material flow status, construct a virtual transmission model in conjunction with the conveyor belt speed, calculate the time node when the material arrives at the drying section, and generate an energy pre-adjustment command to drive the drying equipment to establish a matching thermal field in advance before the material arrives. S3. The stirring behavior of the dried material in the mixing section is monitored in a steady state. By collecting electrical parameters that reflect the changes in stirring load, the rheological characteristic index that quantifies the physical properties of the current batch of material is calculated. S4. Using the generated rheological index as a feedforward input, the rheological index is converted into a physical property correction parameter using a cross-process parameter mapping algorithm, and a power correction command for the microwave sterilization section or a filling position correction command for the molding section is constructed. S5. Real-time assessment of the operational health of each section of the entire line to identify the bottleneck throughput capacity of the control system, using the bottleneck throughput capacity as a global constraint, generating speed modulation commands to drive all field execution units to perform synchronous cycle adjustment. S6. Summarize the generated heat load adjustment data, rheological characteristic index, compensation parameters and velocity change data to generate batch production records containing full-process control logic.

[0016] This invention provides an intelligent control system for large-scale intensive processing of agricultural products. It has the following beneficial effects: 1. This invention constructs a real-time spatiotemporal distribution density data stream of materials through a distributed sensing module, and calculates the time lag using a virtual transmission queue model of a thermal control prediction module. Based on the principle of energy conservation, it generates pre-adjustment commands to drive the heat pump unit, achieving precise feedforward matching of the heat load in the drying section. This effectively overcomes the time lag caused by the physical thermal inertia of the heat pump unit and drying chamber in traditional feedback control, reduces the drying temperature fluctuations caused by feed flow fluctuations, and thus improves the stability and energy efficiency ratio of the drying process.

[0017] 2. This invention utilizes a rheological analysis module to invert and solve the rheological characteristic index representing the physical properties of materials in the steady-state range of the mixing section. Through a collaborative compensation module, compensation instructions for subsequent processes are constructed based on a parameter mapping algorithm. By establishing a cross-process quality feedforward compensation mechanism, the power output of the microwave sterilization section or the filling depth of the forming section can be automatically corrected according to the differences in loose density and flowability of materials between batches. This eliminates the impact of raw material property fluctuations on the physicochemical indicators of the finished product and ensures the consistency of the final product quality.

[0018] 3. This invention calculates the real-time health index based on the statistical variance of state feature data through a cycle modulation module, and identifies the bottleneck throughput capacity of the system based on the short-board effect principle to generate global speed modulation instructions. This achieves adaptive dynamic balance of the entire production cycle, and can automatically adjust the throughput of the entire line to adapt to the bottleneck node when the stability of a single device decreases. This prevents material blockage or unplanned shutdowns caused by single-point performance degradation, and improves the reliability of continuous operation of the production line. Attached Figure Description

[0019] Figure 1 This is a system architecture diagram of the present invention; Figure 2 This is a flowchart of the method steps of the present invention; Figure 3 This is a schematic diagram of the overall control and scheduling module of the present invention; Figure 4 This is a schematic diagram of the thermal control prediction module of the present invention; Figure 5 This is a schematic diagram of the collaborative compensation module of the present invention.

[0020] Among them, 10 is the central control and scheduling module; 20 is the distributed sensing module; 30 is the thermal control prediction module; 40 is the rheological analysis module; 50 is the collaborative compensation module; and 60 is the cycle modulation module. Detailed Implementation

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] See attached document Figure 1 , Figure 1 This is a system architecture diagram according to an embodiment of the present invention. The present invention provides an intelligent control system for large-scale processing of agricultural products. This intelligent control system is based on a distributed control architecture and includes a central control server and multiple field execution units connected via an industrial fieldbus. The field execution units correspond to the pre-processing section, drying section, mixing section, forming section, and packaging section in the production line, respectively.

[0023] The intelligent control system has multiple logic control modules set up in the central control server for performing data processing and command distribution. The intelligent control system includes the following modules: The central control and scheduling module 10 is used to store processing formula data for various agricultural products, and to parse the formula according to the production task to establish data routing between each process node. The distributed sensing module 20 is connected to the sensor located in the cleaning section and is used to collect the spatiotemporal distribution density data and initial temperature data of the material on the conveyor belt. The thermal control prediction module 30 receives data collected by the distributed sensing module 20, which is used to establish a virtual transmission queue model, calculate the time lag of material arrival at the drying section, and generate pre-adjustment instructions for the heat pump unit. The rheological analysis module 40 communicates with the stirring motor driver in the mixing section to collect motor operating status data and calculate the rheological characteristic index of the material. The collaborative compensation module 50 stores a cross-process parameter mapping algorithm, which is used to calculate the microwave sterilization power correction value and the filling position correction value of the molding equipment based on the rheological characteristic index. The beat modulation module 60 is used to monitor the operational health of each key node along the entire line, identify the current bottleneck throughput of the system, and generate synchronous modulation instructions for the transmission speed of each section.

[0024] See attached document Figure 2 , Figure 2 This is a flowchart of method steps according to an embodiment of the present invention. The present invention provides an intelligent control method for large-scale refined processing of agricultural products. Based on an intelligent control system, the method includes the following steps: S1, perform formula loading and route construction. The central control and scheduling module 10 reads the process formula file of the current batch of agricultural products. According to the formula definition, the central control and scheduling module 10 sends instructions to the fieldbus to activate the physical nodes related to the process of the agricultural product, shield irrelevant nodes, and initialize the reference parameters of each sensor and actuator. S2, perform spatiotemporal sensing and drying pre-adjustment. When the material enters the cleaning section, the distribution sensing module 20 generates the distribution data of the material on the conveyor belt through the sensor array; the thermal control prediction module 30 obtains the distribution data and the current transmission speed, and calculates the time lag of the material unit arriving at the drying section; within the preset time window before the material actually arrives at the drying section, the thermal control prediction module 30 sends a frequency conversion adjustment command to the heat pump unit of the drying section to adjust the heat load in the drying chamber; S3, extract intermediate rheological characteristics. When the material enters the mixing section, the rheological analysis module 40 monitors the mixing process. After the mixing enters the steady-state range, the rheological analysis module 40 collects the torque data of the stirring motor and calculates the rheological characteristic index that characterizes the viscosity and density of the material. The rheological characteristic index is transmitted to the collaborative compensation module 50. S4, perform cross-process parameter feedforward compensation. Before the material flows to the subsequent forming or sterilization section, the collaborative compensation module 50 calculates the compensation parameters based on the received rheological characteristic index using a preset mapping algorithm. For the microwave sterilization process, the collaborative compensation module 50 generates a power distribution correction instruction. For the tableting or capsule filling process, the collaborative compensation module 50 generates a filling depth servo correction instruction. The instructions are fed forward to the corresponding field execution unit. S5, perform full-line health state throughput balancing. During the entire production process, the cycle time modulation module 60 periodically collects the operating status data of each process node and calculates the health index of each node. The cycle time modulation module 60 identifies the throughput bottleneck node of the current system based on the health index and calculates the global balanced throughput. Based on the global balanced throughput, the cycle time modulation module 60 sends speed synchronization commands to all involved conveyors and processing equipment along the entire line to adjust the production cycle time. S6, Generate process data records. After the production task is completed, the overall control and scheduling module 10 summarizes the heat load adjustment curve, rheological characteristic index, compensation parameters and speed change curve generated in steps S2 to S5, generates electronic production records for this batch of products and stores them.

[0025] See attached document Figure 3 The central control and scheduling module 10 is deployed in the non-volatile storage medium of the central control server and is executed by the processor. The central control and scheduling module 10 includes a recipe parsing unit, a route construction unit, and a parameter distribution unit.

[0026] The formula parsing unit is equipped with a relational database interface for managing and retrieving standardized process formula data. The process formula data is stored in structured data tables, with each independent agricultural product processing task corresponding to a unique formula identifier. The formula data structure stored in the formula parsing unit... The definition is as follows: ;in, This indicates the product's unique serial number; This is the process topology sequence, used to define the order of physical nodes traversed during the processing flow; This is a set of process parameters, containing the standard operating parameters for each process step. For example, for vegetable powder products, its process topology sequence... Defined as {washing node, blanching node, drying node, pulverizing node, packaging node}; for tablet products, its process topology sequence is... Defined as {mixing node, granulation node, drying node, tableting node, packaging node}, the specific construction method and query statement of the database are well known to those skilled in the art and will not be elaborated here.

[0027] The routing construction unit is used to construct the process topology sequence output by the recipe parsing unit. The routing construction unit establishes data transmission paths and control logic connections between various physical devices by generating a topology control matrix. To manage the enable status of devices on the fieldbus, let the set of system physical nodes be . , where k is the total number of nodes. Topology control matrix for binary vectors:

[0028] in, This represents the enabled state of the i-th physical node, and k represents the total number of physical nodes. When When the routing construction unit sends a closing command to the fieldbus, it activates the corresponding field control unit and its communication port; when At this time, the routing construction unit sends a blocking command to cut off data exchange between the node and the master control server, and locks the physical device corresponding to the node into standby or bypass state. The routing construction unit compares the current task's... Given the system node set N, the S corresponding to nodes not in the sequence i Set to 0. For example, when performing a vegetable powder processing task, the status bit corresponding to the tablet press node is set to 0, thereby preventing erroneous instructions from being sent to irrelevant devices at the logical level.

[0029] The parameter distribution unit is used to distribute process parameter sets The physical quantities in the PLC are mapped to the register address values ​​of specific field execution units. The parameter distribution unit internally stores a device address mapping table, which records the correspondence between the control variables (such as temperature setpoints and motor frequency setpoints) of each physical node and the PLC holding register addresses. After the routing construction unit completes the link establishment, the parameter distribution unit traverses... For each parameter item, the corresponding register address is looked up, and the baseline parameter is written to the corresponding field controller via an industrial Ethernet protocol (such as Modbus TCP or Profinet). The parameter distribution unit is also configured with parameter verification logic. Before distributing parameters, it reads the current status word of the field device to confirm that the device is in "ready" or "remote control" mode to ensure the validity of the parameter writing.

[0030] The distribution sensing module 20 is mainly used in the pre-processing section to convert the continuous material flow into spatiotemporal distribution data that can be used by the control system for prediction calculation. The distribution sensing module 20 includes a data acquisition unit, a density calculation unit, and an initial state output unit.

[0031] The data acquisition unit is equipped with multiple sensor interfaces, including a vision acquisition unit or a weighing sensor array for measuring the linear density distribution of materials, and an infrared temperature sensor for measuring the initial temperature of materials. The vision acquisition unit can use a line scan camera or a high-resolution industrial camera to acquire grayscale image data of the material-covered area on the conveyor belt. The weighing sensor array converts the instantaneous weight signal of a certain section of the conveyor belt into an electrical signal. The data acquisition unit operates at a predetermined sampling frequency. (Hertz) Synchronous acquisition of multiple signals.

[0032] The density calculation unit receives the raw signal from the data acquisition unit and performs preprocessing and density inversion. For visually acquired data, the density calculation unit performs image segmentation, thresholding, and other operations to convert image pixel information into the percentage of material coverage area. Where x represents the position coordinates on the conveyor belt, and the density calculation unit is based on preset material parameters. Calculate the instantaneous linear density of a material using the average mass per unit area. : ; The density calculation unit calculates density by... Integrating within the time window yields the mass flow rate of the material over a given time period. .

[0033] The initial state output unit outputs the instantaneous mass flow rate generated by the density calculation unit. The initial temperature of the material measured by the infrared temperature sensor The combination serves as the key input data for feedforward predictive control. The initial state output unit is equipped with a data forwarding interface for sending real-time data streams to the thermal control prediction module 30. : ; Among them, data stream The thermal control prediction module 30 is used for spatiotemporal lag calculation and feedforward prediction of heat load, where v(t) represents the instantaneous running speed of the conveyor belt.

[0034] See attached document Figure 4 The thermal control prediction module 30 is deployed in the central control server and includes a virtual queue construction unit, a dynamic load calculation unit, and a feedforward execution unit.

[0035] The virtual queue construction unit is used to establish a material transfer model between the outlet of the cleaning section and the inlet of the drying section. The virtual queue construction unit continuously receives the instantaneous running speed v(t) of the conveyor belt and the system's preset physical transfer distance L. For any material unit located at the outlet of the cleaning section at any time t0, the virtual queue construction unit calculates the estimated arrival time of the material unit at the inlet of the drying section through a real-time integration algorithm. The integral relation satisfies: ; Where L represents the preset physical transmission distance, Indicates in The velocity at time t0 represents the initial time, t arr Indicates the estimated time of arrival. The variable is the integral variable. The virtual queue construction unit in the digital control system solves the above integral using a discretized accumulation method, updating the spatiotemporal lag time in real time. ,in The virtual queue building unit maintains a first-in-first-out (FIFO) virtual material queue in memory. Each element in the virtual material queue contains the mass flow rate data of that material unit and its corresponding estimated arrival time t. arr .

[0036] The dynamic load calculation unit is connected to the virtual queue construction unit and is used to calculate future thermal energy demand based on material characteristics. The dynamic load calculation unit reads the virtual material queue corresponding to the future time. Material flow data arriving at the drying section and initial temperature data The dynamic load calculation unit, based on the principle of energy conservation, calculates the load in the drying section. Maintain the target temperature T at all times set Required theoretical thermal power : ; in, Set a constant for the specific heat capacity of materials pre-stored in the formula; The target temperature value set for the drying process; The initial surface temperature of the material as measured by the distributed sensing module; The heat loss power constant of the drying chamber under steady state is determined by the thermal insulation coefficient of the equipment and the ambient temperature difference.

[0037] The feedforward actuator converts theoretical thermal power into specific control commands for the heat pump unit and performs time compensation. Considering the physical response delay between the heat pump compressor and heating components receiving commands and establishing a steady-state thermal field, the feedforward actuator is equipped with a pre-adjustment time window. For the calculated arrival time The feedforward execution unit determines the timing of the adjustment action. : ; At the moment of action At that time, the feedforward execution unit calls the compressor frequency mapping function. Generate frequency control commands for the heat pump unit. : ; in, This indicates a frequency control command for the heat pump unit. Indicates the time of arrival of the action. This represents the theoretical thermal power required at the expected arrival time. The mapping function represents the expected arrival time of the material. It is a nonlinear lookup table or polynomial fitting function based on the performance curve (COP curve) of the heat pump unit, used to map the required heat power to the drive frequency of the variable frequency compressor.

[0038] The feedforward actuator transmits frequency control commands via fieldbus. The field controller sent to the drying section enables the drying chamber to complete the heat load matching before the material actually arrives, eliminating the temperature fluctuation lag in traditional feedback control.

[0039] The rheological analysis module 40 is used to perform online analysis of the physical properties of materials in the mixing section, and to inversely determine the rheological characteristics of the materials by monitoring the driving load characteristics of the mixing equipment. The rheological analysis module 40 includes a steady-state discrimination unit, an electrical parameter acquisition unit, and an exponential calculation unit.

[0040] The steady-state discrimination unit communicates with the mixer's process controller to identify the operational stage of the mixing process and determine an effective data analysis window. During the mixing of powder or granular materials, the load torque tends to stabilize as the mixing uniformity increases. The steady-state discrimination unit has a preset steady-state time interval [t1, t2]. The start time t1 of this interval is set as a specific delay after the mixing starts to avoid the non-steady-state stage of start-up impact and violent material tumbling; the end time t2 is set as the time before the end of the mixing process. The steady-state discrimination unit monitors the mixing timer in real time. When the system time t enters the interval [t1, t2], it generates an enable signal to trigger subsequent data acquisition and calculation processes.

[0041] The electrical parameter acquisition unit establishes a data connection with the frequency converter or servo driver driving the stirring motor. After receiving the enable signal from the steady-state discrimination unit, the electrical parameter acquisition unit synchronously reads the key operating parameters of the stirring motor using a high-frequency sampling method. The operating parameters include the instantaneous drive current of the stirring motor. and instantaneous mechanical angular velocity For variable frequency drive systems using vector control, operating parameters can be read from the internal registers of the drive via an industrial bus; for ordinary motor drive systems, the electrical parameter acquisition unit obtains operating parameters through external Hall current sensors and rotary encoders.

[0042] The index calculation unit receives the data stream from the electrical parameter acquisition unit and performs the calculation of the rheological characteristic index. Based on the physical relationship between the motor output power and the mechanical load torque, the index calculation unit constructs a rheological characteristic calculation model. During the mixing process, the resistance torque experienced by the stirring blades is positively correlated with the mixing viscosity and loose density of the material. The index calculation unit uses the following integral formula to calculate the rheological characteristic index representing the physical properties of this batch of material. : ; In the formula, U is the effective value of the DC bus voltage or stator voltage of the motor driver, which is considered a constant during steady-state operation or is obtained by real-time acquisition; This represents the total duration of the steady-state sampling interval; t1 and t2 are the start and end times of data sampling, respectively. The torque coefficient of the motor is determined by the number of pole pairs and magnetic flux, and is an inherent parameter of the equipment; U represents the effective value of the DC bus voltage or stator voltage of the motor driver. For instantaneous drive current; This refers to the instantaneous angular velocity. The exponential calculation unit eliminates instantaneous fluctuation noise caused by random material flow by averaging the instantaneous torque within the steady-state range, thereby obtaining a stable rheological characteristic exponent. This index serves as a digital fingerprint that quantifies the loose packing density and flowability of the current batch of materials, and is transmitted to the collaborative compensation module 50 for parameter correction in subsequent processes.

[0043] See attached document Figure 5 The collaborative compensation module 50 is used to establish a cross-process feedforward control mechanism between the mixing process and the subsequent finishing process, in order to eliminate product quality fluctuations caused by differences in raw material properties between batches. The collaborative compensation module 50 includes a parameter mapping unit, a power compensation unit, and a position servo unit.

[0044] The parameter mapping unit serves as the input interface for feedforward control, receiving rheological characteristic indices from the rheological analysis module 40. The parameter mapping unit internally stores a multi-physics property database for different material types. This multi-physics property database includes pre-determined rheological characteristic indices and relative permittivity of the materials. The nonlinear mapping relationship curve between them. The parameter mapping unit uses interpolation algorithms or multinomial regression models to map the received rheological characteristic indices. Estimated relative permittivity converted to the current batch of material Relative permittivity It is a physical quantity that represents the ability of a material to absorb waves in an electromagnetic field. Its value is directly affected by the moisture content and density of the material.

[0045] The power compensation unit dynamically corrects the magnetron output power of the downstream microwave sterilization equipment. In the microwave sterilization process, fluctuations in the dielectric constant of the material directly alter the microwave energy absorption efficiency, leading to deviations in sterilization temperature from the set value. The power compensation unit reads the standard rated power set by the system. and the dielectric constant of the standard reference material. The power compensation unit calculates the corrected power setpoint for the microwave magnetron based on the current estimated relative permittivity. The calculation follows the compensation model as follows: ; in, The standard microwave power defined in the process formulation; This is a reference value for the dielectric constant of standard materials at standard moisture content; This is the power compensation gain coefficient, determined by the resonant characteristics of the microwave cavity. Through this compensation model, when the material's moisture content or density is high, leading to an increase in dielectric constant, the system automatically increases the microwave power to ensure a constant energy absorption per unit mass of material, thereby maintaining consistent sterilization results. The power compensation unit will calculate the gain... It is converted into a voltage or duty cycle control signal and fed forward to the power controller of the microwave generator.

[0046] Position servo units are mainly used in solid dosage form production processes, such as tablet presses or capsule filling machines. Position servo units utilize rheological characteristic indices. The ability to characterize the loose density of materials is used to correct the position of the metering mechanism in the molding equipment. In volumetric metering, the weight difference in finished sheets is mainly due to variations in density. The position servo unit acquires the filling depth under standard process conditions. and the corresponding standard rheological index reference value The position servo unit calculates the servo correction value for the filling depth for the filling cam or metering plate. The calculation formula is as follows: ; in, The corrected target location for the filling depth; This refers to the mechanical filling depth corresponding to standard density materials. The rheological index is the benchmark value for standard materials; This is the position correction sensitivity coefficient, used to adjust the correction magnitude. The formula embodies negative feedback logic: when the monitored rheological index... Greater than the benchmark value When this occurs, it indicates that the current material density is too high, and the position servo unit reduces the filling depth accordingly. To maintain the constant quality of individual finished products, the position servo unit sends the corrected depth command to the servo motor driver that drives the metering mechanism. The driver controls the servo motor to rotate a specific angle to adjust the height of the punch or metering plate, achieving micron-level online weight compensation.

[0047] The cycle time modulation module 60 is used to monitor the operating status of all production equipment in real time and dynamically adjust the production throughput based on the current health status of the equipment to achieve adaptive flow balance. The cycle time modulation module 60 includes a health modeling unit, a bottleneck analysis unit, and a synchronization execution unit.

[0048] The health modeling unit collects state characteristic data reflecting the operational stability of equipment. It pre-defines differentiated health evaluation models for different types of physical nodes. Taking the heat-sealing equipment in the packaging section as an example, its key performance indicator is the stability of the sealing pressure. The health modeling unit collects the instantaneous pressure values ​​of the packaging machine cylinders using a sliding window method and calculates the variance of the pressure data within that time window. A larger variance value indicates more unstable equipment operation and a higher risk of failure. The health modeling unit calculates the real-time health index of the node based on the variance. The calculation model is as follows: ;in, This is a normalized health score, with a value range of [value range missing]. , where 1 represents the ideal state of health; The real-time statistical variance of the state parameters; This is a preset sensitivity weighting factor, determined based on the equipment's tolerance to process fluctuations. For other types of equipment, such as tablet presses, the health modeling unit can collect the peak fluctuation rate of the main pressure sensor as the variance input. The degree modeling unit periodically updates all controlled nodes across the entire line. Health Index Set .

[0049] The bottleneck analysis unit receives the index set from the health modeling unit to identify the current system's throughput bottlenecks and calculate the globally optimal production cycle time. The bottleneck analysis unit stores the rated design throughput of each physical node. The bottleneck analysis unit first calculates the maximum allowable throughput that each node can stably operate under the current health status by combining the real-time health index. : ; in, This represents the real-time health index of node i at time t.

[0050] When equipment health deteriorates, the load requirement is proactively reduced to maintain continuous operation and avoid forced shutdown. Based on the principle of the weakest link effect, the bottleneck analysis unit filters the minimum value among the maximum allowable throughput of all nodes to determine the global equilibrium throughput of the entire system. : ; in, Describes the minimum value function; This represents the maximum allowed throughput of the i-th node at the current time; n is the total number of controlled nodes; and t is the current time.

[0051] Global equilibrium flux This indicates the highest overall output speed that can be achieved at the current moment to ensure that all equipment on the entire line does not experience overload shutdown.

[0052] The synchronous execution unit is used to convert the calculated global balanced throughput into speed control commands for each specific actuator. The synchronous execution unit is equipped with a process transfer ratio database for all equipment on the production line, which records the linear or nonlinear proportional relationships between each conveyor motor, machining host and system throughput. The synchronous execution unit calculates the target running speed for each controlled drive node j. : ; in, The process transfer ratio coefficient for the j-th device is used. The synchronous execution unit transmits the calculated speed command synchronously to all frequency converters and servo drives in the entire line via industrial fieldbus, using broadcast or multicast. Upon receiving the new speed command, each drive smoothly adjusts the motor speed according to the preset acceleration / deceleration slope. When a single device in the system (such as a packaging machine) experiences a status fluctuation that affects its health... During descent, the cycle modulation module 60 automatically reduces the overall operating speed of the line, causing upstream processes such as cleaning, drying, and mixing to slow down synchronously, thereby preventing material from accumulating at bottleneck nodes and realizing adaptive flexible production of the production line under non-ideal operating conditions.

[0053] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent control system for large-scale precision processing of agricultural products, characterized in that, Comprise: A general control scheduling module for parsing the processing recipe data of agricultural products to extract the process topology sequence, and selectively activating the corresponding physical nodes based on the definition of the process topology sequence to build data routing connections between process nodes; A distribution perception module for obtaining material detection signals of the pre-processing section, inversely calculating the instantaneous linear density of the material based on the material detection signals, and combining the initial temperature of the material to build real-time spatio-temporal distribution density data flow of the material flow state on the conveyor belt; A thermal control prediction module for building a virtual transmission queue model using the real-time spatio-temporal distribution density data flow and the conveyor belt running speed, calculating the time lag of the material unit arriving at the drying section, and generating a pre-adjustment instruction for driving the heat pump unit before the material unit arrives based on energy demand prediction; A rheological analysis module for monitoring the stirring mixing process of the mixing section, and inversely calculating the rheological characteristic index representing the physical properties of the dried material using electrical parameters reflecting the stirring load characteristics; A collaborative compensation module for mapping the rheological characteristic index to a physical property correction parameter, and generating a power correction value for the microwave sterilization section or a filling position correction value for the forming section based on the physical property correction parameter; A beat modulation module for calculating real-time health degree based on operating state data, identifying the bottleneck flux capacity of the control system according to the real-time health degree, and converting it into speed modulation instructions adapted to each execution unit.

2. The intelligent control system for large-scale precision processing of agricultural products according to claim 1, characterized in that, The general control scheduling module comprises: A recipe analysis unit for outputting recipe data containing process topology sequence; A routing construction unit for generating a topology control matrix in the form of a binary vector according to the process topology sequence, sending a closing instruction to activate the corresponding node to the field execution unit in the process topology sequence by comparing the process topology sequence of the current task with the control system node set, and sending a shielding instruction to lock in standby state to the field execution unit not in the process topology sequence.

3. The intelligent control system for large-scale precision processing of agricultural products according to claim 1, characterized in that, The distribution perception module comprises: A density calculation unit for pre-processing image data or weight signals in the material detection signal to obtain the coverage area or instantaneous weight of the material, calculating the instantaneous linear density based on the average mass per unit area, and generating an instantaneous mass flow as a feedforward prediction input by cumulatively operating the product of the instantaneous linear density and the instantaneous running speed of the conveyor belt within a time window.

4. The intelligent control system for large-scale precision processing of agricultural products according to claim 1, characterized in that, The thermal control prediction module comprises: A virtual queue construction unit for maintaining a virtual transmission queue, and updating the expected arrival time of the material unit arriving at the drying section in real time by discretely accumulating the instantaneous running speed of the conveyor belt; A dynamic load calculation unit for calculating the theoretical heat power required to maintain the target temperature of the drying section based on the principle of energy conservation; A feedforward execution unit for calculating the action trigger time earlier than the expected arrival time based on a preset response time window, and converting the theoretical heat power into a frequency control instruction for driving the drying section heat pump unit when the action trigger time is reached.

5. The intelligent control system for large-scale precision processing of agricultural products according to claim 1, characterized in that, The rheological analysis module comprises: A steady state discrimination unit for determining whether the mixing time enters a preset steady state time interval; An exponential operation unit is configured to collect the instantaneous driving current, instantaneous mechanical angular velocity and voltage data of the stirring motor synchronously in a steady state time interval, calculate the ratio of the stirring motor output power to the mechanical angular velocity, and perform time average processing in the steady state interval to solve the rheological characteristic index for quantitatively representing the bulk density and flowability of the material.

6. The intelligent control system for large-scale precision processing of agricultural products according to claim 1, characterized in that, The cooperative compensation module comprises: A parameter mapping unit is configured to convert the received rheological characteristic index into an estimated relative dielectric constant by using a pre-stored nonlinear mapping relationship curve; A power compensation unit is configured to adjust the standard microwave power according to the deviation proportion of the estimated relative dielectric constant relative to the dielectric constant of the standard reference material, and generate a corrected power setting value for the microwave magnetron according to the constant energy absorption principle.

7. The intelligent control system for large-scale precision processing of agricultural products according to claim 6, characterized in that, The cooperative compensation module further comprises: A position servo unit is configured to represent the material density change by the difference between the rheological characteristic index and the standard rheological index reference value, calculate a filling depth servo correction value for the metering mechanism of the forming section based on negative feedback logic, and drive the servo motor to reduce the filling depth when the rheological characteristic index indicates that the material density is greater than the reference value, so as to keep the weight of the single product constant.

8. The intelligent control system for large-scale precision processing of agricultural products according to claim 1, characterized in that, The beat modulation module comprises: A health degree modeling unit is configured to collect state characteristic data reflecting the running stability of the equipment in a sliding window manner, calculate the statistical variance of the state characteristic data in the current sliding window, and generate a normalized real-time health degree index based on the statistical variance, wherein the greater the statistical variance, the lower the value of the generated real-time health degree index.

9. The intelligent control system for large-scale precision processing of agricultural products according to claim 8, characterized in that, The beat modulation module further comprises: A bottleneck analysis unit is configured to calculate the product of the rated design flux of each section and the real-time health degree index to obtain the maximum allowed flux, and select the minimum value of the maximum allowed flux of all sections as the global balanced flux based on the short board effect principle; A synchronous execution unit is configured to convert the global balanced flux into a target running speed instruction for each execution unit according to a pre-set process transmission ratio coefficient.

10. An intelligent control method for large-scale precision processing of agricultural products, applied to the intelligent control system for large-scale precision processing of agricultural products according to any one of claims 1-9, characterized in that, The method comprises the following steps: S1, analyzing the processing formula data to lock the target process topology sequence, and publishing control instructions to the field bus according to the definition logic of the process topology sequence to activate specific field execution units, and building a bottom-layer data interaction channel throughout the whole line; S2, obtaining the material detection signal of the pre-processing section, converting it into real-time spatiotemporal distribution density data stream reflecting the material flow state, combining with the conveying belt speed to build a virtual transmission model, calculating the time node of the material reaching the drying section, and generating an energy pre-adjustment instruction to drive the drying equipment to establish a matching thermal field in advance before the material arrives; S3, performing steady-state monitoring on the stirring behavior of the material after drying in the mixing section, and inversely solving the rheological characteristic index quantitatively representing the physical properties of the current batch of material by collecting electrical parameters reflecting the stirring load change; S4, using the generated rheological characteristic index as a feedforward input, converting the rheological characteristic index into a physical property correction parameter by using a cross-process parameter mapping algorithm, and building a power correction instruction for the microwave sterilization section or a filling position correction instruction for the forming section; S5, real-time evaluation of the operation health degree of each section of the whole line to identify the bottleneck flux capacity of the control system, taking the bottleneck flux capacity as a global constraint condition, generating a speed modulation instruction to drive all field execution units to execute synchronous beat adjustment; S6, aggregate the generated heat load adjustment data, flow behavior characteristic index, compensation parameter and speed change data, and generate a batch production record containing the whole process control logic.

Images